Kahawai FAP

Kahawai Final Advice Paper

Ministry of Fisheries

29 June 2004

43. The MFish initial view of the 1996 stock assessment was summarised in the IPP at paragraphs 116-124. In summary, a 1996 stock reduction model was used to obtain estimates of virgin and current biomasses and maximum constant yield (MCY) for a single nationwide kahawai stock. Recruitment was assumed to be deterministic for biomass estimation, but stochastic for yield estimation. Estimates of MCY were calculated for a single national fishstock using a model constrained to ensure that the biomass did not go below 20% of the unfished (or virgin) biomass (B0) more than 10% of the time.
44. In the absence of information specific to kahawai, a number of parameter values used in the model were assumed or input as ranges. The natural mortality was one such parameter. In the IPP MFish proposed that a natural mortality of 0.2 for kahawai was the best available information and accordingly proposed that the historic MCY estimates based on that value were the best available. This was because analysis suggested the natural mortality for kahawai is unlikely to be higher than 0.2 and is likely to be close to this estimate.
45. Estimates of MCY derived from the 1996 model were generally regarded as conservative because some of the other parameters used as model inputs were based on maximum observed values. However, in the IPP MFish noted that the historic stock assessment contains important uncertainties, most notably the value used to reflect fishing mortality and the non-commercial catch history applied to the model.
46. A base case model using a 1996 estimate of recreational catch provided an MCY estimate of 7,600 tonnes (MCYbase case). Using different assumptions concerning the pattern and history of non-commercial catches of kahawai, a sensitivity analysis was undertaken where the non-commercial catch was extended back in time and was greater between 1945 and 1992 than the recreational catch used in the base case. This had the effect of increasing estimates of B0, and MCY and was the basis for the greater MCY estimate of 8,200 tonnes (MCYsensitivity analysis).
47. MFish noted in the IPP at paragraph 24 that combined estimates of non-commercial and commercial utilisation for kahawai stocks were just within the estimates for MCY (7 600 and 8 200 tonnes) based on a natural mortality (M) of 0.2. MFish proposed setting combined TACs for fishstocks that coincidently were consistent with the smaller of the two 1996 estimates of sustainable yield (ie 7 600 tonnes). The estimated levels of utilisation for all sectors combined could be accommodated within the proposed TACs and MFish therefore suggested there was no scarcity within the fishery and therefore no clear-cut requirement to consider reducing the current catch in the fishery. MFish noted that the initial proposals were based on the assumption that the stock is at or above BMSY

48. Sanford submits that the methodology proposed for setting TACs for kahawai in the IPP is flawed and fails to properly apply the provisions of the Act. It submits that TACs should be based on the use of best available information. It submits that the 1996 stock assessment is better information as it is based on an evaluation of the sustainability of the resource. Accordingly, Sanford proposes that TACs are set on the basis of the alternative MCY (8 200 tonnes).
49. Sanford notes that the stock assessment estimated that biomass in the mid-1990s was around 50% of virgin biomass (B0), well above BMSY (the biomass that provides the maximum sustainable yield), indicating a healthy kahawai resource at that time. It notes that the IPP conservatively estimated MCY to range between 7 600 and 8 200 tonnes.
50. It submits that the stock assessment, completed almost 20 years after the commencement of the kahawai purse seine fishery, indicated that the stock was being fished at conservative levels. Sanford notes that commercial catches over the decade leading up to the assessment averaged around 6 000 tonnes, while catches in the last decade have averaged less than 4 500 tonnes. It believes that the 1996 stock assessment remains the best available information and should be used as the basis for setting TACs. It submits support for combined TACs based on 8,200 tonnes on the basis that MCY sensitivity analysis was considered to be conservative.
51. Non-Commercial Fishers submit that it is not convinced that the critical values used in the 1996 stock assessment are correct. It submits that any TACs set should be based on the use of the best available information (for estimating MCY) and submits that this fishery should be managed above the biomass that will support maximum sustainable yield.
52. Non-Commercial Fishers note that the IPP preference is for the MCYbase case of 7 600 tonnes. Further, the submission notes that MCYbase case is based on a natural mortality of 0.20 and a model that allows fishing stocks down to a level of biomass that is less than 20% of virgin biomass. It submits that TACs should be set on the basis of the best available estimate of natural mortality, which it considers to be 0.18, and in addition must set out to manage this fishery above the biomass that will support MSY. It submits that recalculating MCY on the basis of M = 0.18 will achieve this. Non-Commercial Fishers submitted a revised estimate of MCY of 6 900 tonnes and recommended combined TACs for all fishstocks be based on this estimate.
53. The RFC submits that it does not consider that a reliable stock assessment can be carried out without a recruitment index for the fishery.
    

54. MFish notes that both commercial and some recreational submissions support the use of the 1996 MCY estimates as a basis for setting TACs. However, submitters differ on which of the MCY estimate should be used. Sanford supports combined TACs for all fishstocks of 8 200 tonnes on the basis that MCYsensitivity analysis estimates were considered to be conservative and constitute the best available information. Non-Commercial Fishers recommend combined TACs for all fishstocks of 6 900 tonnes based on “revising” the MCYbase case estimate using their preferred estimate of natural mortality (M=0.18), which they believe to be a key parameter in the model.
55. The historic stock assessment model used to estimate MCYbase case was based on 1996 recreational harvest. Expert advice is now that the 1996 estimates of recreational catch are unreliable and should not be used. This is because of methodological problems with the 1996 survey.
56. It is possible that the 1996 survey under-estimated recreational catch but this is not certain. If this were the case then the effect on MCY estimates would be to increase them. This is demonstrated by the value of MCY sensitivity analysis (8 200 tonnes), which was determined from the model using higher values of recreational catch than those used in the base case model.
57. On the other hand, MFish notes that adoption of the Non-Commercial Fishers suggestion of using M=0.18 would alter the M=0.2 estimate of MCYsensitivity analysis (8 200) down to approximately 7 600. For MCYbase case the reduction using M=0.18 would be from 7 600 down to approximately 6 600.
58. MFish accepts that M=0.18 may be a more appropriate value for kahawai and notes that while it was not used in the historic assessment model (a range of values was used instead) it is reported in the Stock Assessment Plenary Report as the applicable value for kahawai. However, MFish notes that if the 1996 recreational catch was under-estimated, this counter balances the altered parameter for natural mortality to a degree. The effect of an ad hoc revision of the two parameters in response to stakeholder submissions leads MFish to conclude that a conservative estimate of MCY remains approximately 7 600 tonnes.
59. The simplistic historic assessment remains a reference point for a level of yield from the kahawai fishery. However, you should note that there is considerable uncertainty associated with the historic stock assessment and the resulting MCY estimates. It is also important to note that MCY will only maintain the stock at or above BMSY if it is at or near this level already. If it is substantially lower then lower catch levels may be required to rebuild the stock.
60. MFish agrees with the RFC submission that recruitment variability is a potentially important factor that is poorly known. The 1996 assessment ran a broad range of recruitment sensitivities and selected 0.6 as a conservative value (high variability resulting in lower MCY estimates). The 1996 assessment report noted that recruitment variability may be high for kahawai and the establishment of a recruitment index would give one means of improving the biomass estimates. Attempts to establish a recruitment index for kahawai to date have not been successful.
61. As noted in this paper and in the IPP, there is considerable uncertainty regarding the historic assessment, which is now six years out of date. A new assessment of the kahawai stock is required.
62. The historic assessment utilised a basic modelling approach and did not incorporate any abundance index. New information is available that would allow further evaluation of critical parameters incorporated in the historic assessment (e.g. revised catch histories and revised estimates of natural mortality). Tag data available at the time of the historic assessment could not be incorporated into that assessment.
63. Additional research has occurred since the historic assessment providing additional information that could be incorporated into a new assessment approach (e.g. catch-at-age sampling). Research is currently underway investigating one relative index of kahawai abundance and an index of recruitment (refer RFC submission). Additional research is soon to be contracted to investigate another relative index of kahawai abundance.
64. New assessment methods are available that can better utilise all of the available data, whether a relative index or indices are successfully developed or not, and a new assessment is therefore to be proposed for the 2005-06 year (with the historic approach repeated for comparative purposes only). Information from this assessment should be available for reviewing management arrangements for the 2006-07 fishing year.